This commit creates file structure for register cache related
functions.
Specifically:
* `riscv_reg.h` -- general interface to registers. Safe to use after
register cache initialization is successful.
* `riscv_reg_impl.h` -- helper functions to use while implementing
register cache initialization.
* `riscv_reg.c` -- definitions of functions from `riscv_reg.h` and
`riscv_reg_impl.h`.
* `riscv-011_reg.h` -- register cache interface specific to 0.11
targets.
* `riscv-013_reg.h` -- register cache interface specific to 0.13+
targets.
* `riscv-011/0.13.h` -- version-specific methods used to access
registers. Will be extended as needed once other functionality (not
related to register access) is separated (e.g. DM/DTM specific stuff).
Change-Id: I7918f78d0d79b97188c5703efd0296660e529f2a
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
When an asynchronous exception occurs at the same time
as a breakpoint event (either hardware breakpoint or software breakpoint),
it is possible for the processor to halt at the beginning of the
exception handler instead of the instruction address pointed
by the breakpoint.
During debug entry in exception handler state and with BKPT bit set
as the only break reason in DFSR, check if there is a breakpoint, which
have triggered the debug halt. If there is no such breakpoint,
resume execution. The processor services the interrupt and
halts again at the correct breakpoint address.
The workaround is not needed during target algo run (debug_execution)
because interrupts are disabled in PRIMASK register.
Also after single step the workaround resume never takes place:
the situation is treated as error.
Link: https://developer.arm.com/documentation/SDEN1068427/latest/
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I8b23f39cedd7dccabe7e7066d616fb972b69f769
Reviewed-on: https://review.openocd.org/c/openocd/+/8332
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Liviu Ionescu
Conflicts:
* `doc/openocd.texi`: due to d382c95d57,
resolved by selecting the upstream version.
* `src/server/gdb_server.c`: between
944fe66f10 and
92e8823ebd. Resolved by adopting the use
of `LOG_TARGET_*`.
* `src/target/target.c`: between
639e68a621 and
c5358c84ad, selected the version from
`riscv-openocd`.
Change-Id: Ic1327f25e147945e0ec82947a82452501e8ee5de
Most of the work is already done by [1].
Remove few more '_s' suffix and also fix some comment referring to
the old name of the struct.
Link: https://review.openocd.org/c/openocd/+/8340
Change-Id: Ifddc401c3b05e62ece3aa7926af1e78f0c4a671e
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8341
Reviewed-by: zapb <dev@zapb.de>
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
The output "gdb port disabled" is confusing without reference to the
target. Use LOG_TARGET_INFO() to output the target name.
While at it, use LOG_TARGET_xxx() for all log statements where the
target name is already used.
Change-Id: I70b134145837db623e008a4a6c0be0008d9a0d87
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8313
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Changes affect the function remote_bitbang_fill_buf.
When read_socket returns 0, socket reached EOF and there is
no data to read. But if request was blocking, the caller
expected some data. Such situations should be treated as ERROR.
Change-Id: I02ed484e61fb776c1625f6e36ab14c85891939b2
Signed-off-by: Timur Golubovich <timur.golubovich@syntacore.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8325
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Commit f9509c92db ("itm: rework itm commands before 'init'")
ignores the default enable of ITM channel 0, that is applied when
no 'itm port[s]' is issued.
Call armv7m_trace_itm_config() unconditionally to handle it.
Change-Id: I3e85d0b063ed38c1552f6af9ea9eea2e76aa9025
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reported-by: Paul Fertser <fercerpav@gmail.com>
Fixes: f9509c92db ("itm: rework itm commands before 'init'")
Reviewed-on: https://review.openocd.org/c/openocd/+/7900
Reviewed-by: <post@frankplowman.com>
Tested-by: jenkins
The register SPSR_EL1 is accessible and it's content is relevant
only when the target is in EL1 or EL2 or EL3.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd5384000, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $SPSR_EL1
or through OpenOCD command
reg SPSR_EL1
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Change-Id: Ia0f984d52920cc32b8ee31157d62c13dea616a3a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8276
Tested-by: jenkins
The register ESR_EL1 is accessible and it's content is relevant
only when the target is in EL1 or EL2 or EL3.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd5385200, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ESR_EL1
or through OpenOCD command
reg ESR_EL1
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Change-Id: Icd65470c279e5cfd03091db6435cdaa1c447644c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8275
Tested-by: jenkins
The register ELR_EL1 is accessible and it's content is relevant
only when the target is in EL1 or EL2 or EL3.
Without this patch, an error:
Error: Opcode 0xd5384020, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ELR_EL1
or through OpenOCD command
reg ELR_EL1
Detect the EL and return error if the register cannot be accessed.
Change-Id: I402dda4cd9dae502b05572fc6c1a8f0edf349bb1
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8274
Tested-by: jenkins
The register SPSR_EL2 is accessible and it's content is relevant
only when the target is in EL2 or EL3.
Virtualization SW in EL1 can also access it, but this either
triggers a trap to EL2 or returns SPSR_EL1. Debugger should not
mix the real SPSR_EL2 with the virtual register.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd53c4000, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $SPSR_EL2
or through OpenOCD command
reg SPSR_EL2
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Change-Id: If3792296b36282c08d597dd46cfe044d6b8288ea
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8273
Tested-by: jenkins
The register ESR_EL2 is accessible and it's content is relevant
only when the target is in EL2 or EL3.
Virtualization SW in EL1 can also access it, but this either
triggers a trap to EL2 or returns ESR_EL1. Debugger should not mix
the real ESR_EL2 with the virtual register.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd53c5200, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ESR_EL2
or through OpenOCD command
reg ESR_EL2
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Change-Id: Icb32b44886d50907f29b068ce61e4be8bed10208
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8272
Tested-by: jenkins
The register ELR_EL2 is accessible and it's content is relevant
only when the target is in EL2 or EL3.
Virtualization SW in EL1 can also access it, but this either
triggers a trap to EL2 or returns ELR_EL1. Debugger should not mix
the real ELR_EL2 with the virtual register.
Without this patch, an error:
Error: Opcode 0xd53c4020, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ELR_EL2
or through OpenOCD command
reg ELR_EL2
Detect the EL and return error if the register cannot be accessed.
Change-Id: Idf02b42a7339df83260c1e44ceabbb05fbf392b9
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8271
Tested-by: jenkins
The register SPSR_EL3 is accessible and it's content is relevant
only when the target is in EL3.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd53e4000, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $SPSR_EL3
or through OpenOCD command
reg SPSR_EL3
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Change-Id: I00849d99feeb96589c426fcafda98127dbd19a67
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8270
Tested-by: jenkins
The register ESR_EL3 is accessible and it's content is relevant
only when the target is in EL3.
Plus, the register is 64 bits wide.
Without this patch, an error:
Error: Opcode 0xd53e5200, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ESR_EL3
or through OpenOCD command
reg ESR_EL3
Detect the EL and return error if the register cannot be accessed.
Handle the register as 64 bits.
Drop the FIXME comment on Aarch32 case, as the register exists in
Aarch64 only.
Change-Id: Ie8c69dc7b50ae81a52506cf151c8e64e15752d0d
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8269
Tested-by: jenkins
The register ELR_EL3 is accessible and it's content is relevant
only when the target is in EL3.
Without this patch, an error:
Error: Opcode 0xd53e4020, DSCR.ERR=1, DSCR.EL=1
is triggered by GDB register window or through GDB command
x/p $ELR_EL3
or through OpenOCD command
reg ELR_EL3
Detect the EL and return error if the register cannot be accessed.
Change-Id: I545abb196e5c34e462c7e5d5d3ec952e588642da
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8268
Tested-by: jenkins
The command 'gdb_report_register_access_error' is used to silence
errors while reading registers and not reporting them to GDB.
Nevertheless, the error is printed by a LOG_ERROR() in armv8_dpm.
Change the message to LOG_DEBUG().
It will still cause the error to be propagated and eventually
printed by the caller (e.g. by the command 'reg').
Change-Id: Ic0db74fa28235d686ddd21a5960c52ae003e0931
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8267
Tested-by: jenkins
These functions are today always called with non-NULL parameter
regval, so the actual check is not needed.
Anyway, for any future code change, check the parameter at the
entry of the functions and return error if it is not valid.
Simplify the check to assign the result value and align the code
of the two functions.
Change-Id: Ie4d98063006d70d9e2bcfc00bc930133caf33515
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8266
Tested-by: jenkins
Use LOG_TARGET_ERROR() to print the error messages and additionally add
a reference to the related target.
Change-Id: I06722f3911ef4034fdd05dc9b0e2571b01b657a4
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8314
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
cortex_m_poll_one() detects reset testing S_RESET_ST sticky bit.
If the signal comes unexpectedly, poll must return TARGET_RESET state.
On the contrary in case of polling inside of an OpenOCD reset command,
TARGET_RESET has been has already been set and we need to get out of
it as quickly as possible.
The original code needs 2 polls: the first clears S_RESET_ST
and keeps TARGET_RESET state, the current TARGET_RUNNING or TARGET_HALTED
is reflected as late as the second poll is done.
Change the logic to keep in TARGET_RESET only when necessary.
See also [1]
Link: [1] 8284: tcl/target: ti_cc3220sf: Use halt for CC3320SF targets | https://review.openocd.org/c/openocd/+/8284
Fixes: https://sourceforge.net/p/openocd/tickets/360/
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I759461e5f89ca48a6e16e4b4101570260421dba1
Reviewed-on: https://review.openocd.org/c/openocd/+/8285
Tested-by: jenkins
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Remove list of id codes for all families.
Maintain a list with id, bscan-length and check position
in the tcl config files for each family.
The Intel FPGA Driver option 'family' is not otional anymore.
Change-Id: I9a40a041069e84f6b4728f2cd715756a36759c89
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/8083
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
In case of fail to allocate 'obj->name', the memory allocated for
'obj->out_filename' is not freed, thus leaking.
Since 'obj' is allocated with calloc(), thus zeroed, switch to use
the common error exit path for both allocations of 'obj->name' and
'obj->out_filename'.
Fixes: 2506ccb509 ("target/arm_tpiu_swo: Fix division by zero")
Change-Id: I412f66ddd7bf7d260cee495324058482b26ff0c5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8300
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Without the selection the TAP can be left in bypass.
Change-Id: I79c6bf74802dc9c9475947d1787a3d0b797f3952
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Erase is initiated by write to a flash address. Due to the
silicon errata of nRF91 the write stalls the bus until the page erase
is finished (takes up to 87ms).
If the adapter does not handle SWD WAIT properly, the following read
in nrf5_wait_for_nvmc() returns ERROR_WAIT.
Wait for fixed time before accessing AP. Not nice, but the only
working solution until all adapters handle SWD WAIT.
If the fixed wait does not suffice, continue the wait loop after a delay.
It makes some unnecessary noise however erase works.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I63faf38dad79440a0117ed79930442bd2843c6db
Reviewed-on: https://review.openocd.org/c/openocd/+/8115
Reviewed-by: Tomáš Beneš <tomas@dronetag.cz>
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Since nRF9160 Product Specification v2.1 the new UICR SIPINFO
fields should be preferred over UICR INFO.
Tested on nRF9161.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: Ib8005b3b6292aa20fa83c1dcebd2de27df58b661
Reviewed-on: https://review.openocd.org/c/openocd/+/8114
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Probes all flash and UICR areas.
Flash erase and write tested.
On nRF53 mass erase works on the application core flash bank only.
The Tcl script nrf53_recover can serve as the workaround on the
network core.
TODO: mass erase of the nRF53 network core flash.
Some ideas taken from [1] and [2].
Change-Id: I8e27a780f4d82bcabf029f79b87ac46cf6a531c7
Link: [1] 7404: flash: nor: add support for Nordic nRF9160 | https://review.openocd.org/c/openocd/+/7404
Link: [2] 8062: flash: nor: add support for Nordic nRF9160 | https://review.openocd.org/c/openocd/+/8062
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8112
Reviewed-by: Tomáš Beneš <tomas@dronetag.cz>
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Enable and disable erase mode only once
instead of toggling it for each sector.
Refactor to decrease the number of call levels.
Change-Id: Ie546a4fc24da0eea2753a2bebaa63d941ef7aa1d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8111
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Preparatory change before extending support to nRF53 and 91.
While on it, rename nRF51 and 52 specific routines and constants.
Change-Id: I46bc496cef5cbde46d6755a4b908c875351f6612
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8110
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
For configurations which include multiple targets and the "pipe" mode is
requested only the first gdb_server instance should be enabled,
otherwise GDB gets confusing replies, goes out of sync and the session
fails in weird ways.
Compile-tested only.
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Change-Id: If8f13aa7b58e9b0dc6d5ae88cf75538b34cc1218
Reviewed-on: https://review.openocd.org/c/openocd/+/8222
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
this functionality allows to query if a target belongs to some smp group
and to dynamically turn on/off smp-specific behavior
Change-Id: I67bafb1817c621a38ae4a2f55e12e4143e992c4e
Signed-off-by: Parshintsev Anatoly <anatoly.parshintsev@syntacore.com>
Signed-off-by: Bernhard Rosenkränzer <bero@baylibre.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8296
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
LOG_USER only outputs to user interfaces, but leaves no way to get the
FTDI inputs over the RPC interface. Switch to command_print so this
string goes to both logs and the RPC interface.
Change-Id: I99024194b6687b88d354ef278aa25f372c862c22
Signed-off-by: Mark Featherston <mark@embeddedts.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8294
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
For some target, the API assert_reset() checks if the target has
been examined, with target_was_examined(), to perform conditional
operations like:
- assert adapter's srst;
- write some register to catch the reset vector;
- invalidate the register cache.
Targets created with -defer-examine gets the examine flag reset
right before entering in their assert_reset(), disrupting the
actions above.
For targets created with -defer-examine, move the reset examine
after the assert_reset().
Change-Id: If96e7876dcace8905165115292deb93a3e45cb36
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8293
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Commit 236c54c94a ("server/gdb_server.c: support unavailable
registers") correctly returns a string of 'x' when the register is
not available in the current target.
While implementing this, it incorrectly drops the pre-existing
feature of optionally ignoring errors while reading a register.
This feature has a real use case documented in the OpenOCD manual
in chapter 'Using GDB as a non-intrusive memory inspector', where
GDB attaches to a target without halting it. For targets that need
to be halted to read its registers, we need to hack the values of
the registers returned to GDB; either returning 'xxxx' or an error
causes GDB to drop the connection.
Re-add the check on 'gdb_report_register_access_error' to keep the
pre-existing behavior when a register error has to be ignored:
- return a string of '0';
- drop a debug message.
Change-Id: Ie65c92f259f92502e688914f334655b635874179
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: 236c54c94a ("server/gdb_server.c: support unavailable registers")
Reviewed-on: https://review.openocd.org/c/openocd/+/8228
Tested-by: jenkins
This allows to eliminate up to two DMI NOPs.
Change-Id: I09a18bd896fce2392d1b65d4efb38b53e334a358
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Caching is somewhat handled in `riscv-011.c`. Handling it additionaly in
`riscv.c` may cause problems. Sice there is no simulator that supports
RISC-V Debug Specification v0.11, so it is not feaseable to automate
testing.
This commit separates 0.11 register accesses and unlocks further
development in this area.
Change-Id: I73ff17ef85106c4ababa38319f446f6c384a1750
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
This reverts commit 9d4df3420c.
I believe the reasoning behind this workaround is no longer valid.
Change-Id: Ie8705f75eb8ad7b72fc8ffcf39125be764cb43be
Code cleanup: "slot_t" is unused in riscv013 - remove it.
Change-Id: I9d5a0cf8446a180b1d13a9ce2c86d904b946cf28
Signed-off-by: Jan Matyas <jan.matyas@codasip.com>
Conflict in src/rtos/FreeRTOS.c due to
fbea7d5d38 -- resolved by replacing
`target->type->name` with a call to `target_type_name()`.
Change-Id: I56702c6133894458903de7a4d764903004aa8b86
hide_csrs should not emit warnings on an attempt to hide non-exitents CSR.
hide_csrs funcitonality is intended to be used for scenarios when we don`t
want certain groups of registers to be available in GDB. Typically this is
needed to simplify integration with various IDE. In such scenarious it may
be impractical/unfeseable to figure out which register is present on a
target. So reporting a situation when a user wants to hide a non-existent
register creates way too much noise. This commit reduces severity of
relevant debug message to LOG_TARGET_DEBUG
Before this patch the following behavior is observed on targets that do
not support hit bit:
```
bp 0x80000004 4 hw
resume 0x80000000
riscv.cpu halted due to watchpoint
```
This happens because the current implementation relies on the presence
of hit bit way too much. While working on this patch few defects in
hit bit-based trigger detection were discovered, added appropriate
TODOs.
Few files include target_type.h even if it is not needed.
Drop the include.
Other files access directly to target type's name instead of using
the proper API target_type_name().
Use the API and drop the include.
Change-Id: I86c0e0bbad51db93500c0efa27b7d6f1a67a02c2
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8260
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
clang doesn't support the gnu_printf attribute that OpenOCD uses if
__GNUC__ and __GNUC_MINOR__ indicates gcc >= 4.4.
Most clang builds set __GNUC__/__GNUC_MINOR__ to 4.2 for historical
reasons, so they don't trigger this condition; however, some builds set
it to something much higher to work around code using __GNUC__ to
determine if a feature that does exist in clang (but not gcc 4.2) is
available, causing OpenOCD to use attribute gnu_printf.
The problem can be reproduced without a special clang build by adding
-fgnuc-version=14.1 to CFLAGS.
Change-Id: I3c0832d4201578b116c5214203f95b6153dad30e
Signed-off-by: Bernhard Rosenkränzer <bero@baylibre.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8258
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Do not free the service in 'connection_closed_handler' because it is
free'd by the server infrastructure.
Checkpatch-ignore: COMMIT_LOG_LONG_LINE
This error was detected with valgrind:
==272468== Invalid free() / delete / delete[] / realloc()
==272468== at 0x484B27F: free (in /usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so)
==272468== by 0x1F34C7: remove_service (server.c:374)
==272468== by 0x2ED3D5: semihosting_tcp_close_cnx (semihosting_common.c:1819)
==272468== by 0x2ED3D5: handle_common_semihosting_redirect_command (semihosting_common.c:1926)
==272468== by 0x1FC703: exec_command (command.c:520)
==272468== by 0x1FC703: jim_command_dispatch (command.c:931)
==272468== by 0x36980F: JimInvokeCommand (in /home/marc/openocd/build/src/openocd)
==272468== by 0x1FFFFFFFFF: ???
==272468== by 0x53ED09F: ???
==272468== by 0x300000001: ???
==272468== by 0x1FFEFFF7FF: ???
==272468== by 0x3D3984: ??? (in /home/marc/openocd/build/src/openocd)
==272468== by 0x2: ???
==272468== Address 0x5fff650 is 0 bytes inside a block of size 24 free'd
==272468== at 0x484B27F: free (in /usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so)
==272468== by 0x2ECA42: semihosting_service_connection_closed_handler (semihosting_common.c:1807)
==272468== by 0x1F2E39: remove_connection.isra.0 (server.c:164)
==272468== by 0x1F349E: remove_connections (server.c:350)
==272468== by 0x1F349E: remove_service (server.c:364)
==272468== by 0x2ED3D5: semihosting_tcp_close_cnx (semihosting_common.c:1819)
==272468== by 0x2ED3D5: handle_common_semihosting_redirect_command (semihosting_common.c:1926)
==272468== by 0x1FC703: exec_command (command.c:520)
==272468== by 0x1FC703: jim_command_dispatch (command.c:931)
==272468== by 0x36980F: JimInvokeCommand (in /home/marc/openocd/build/src/openocd)
==272468== by 0x1FFFFFFFFF: ???
==272468== by 0x53ED09F: ???
==272468== by 0x300000001: ???
==272468== by 0x1FFEFFF7FF: ???
==272468== by 0x3D3984: ??? (in /home/marc/openocd/build/src/openocd)
==272468== Block was alloc'd at
==272468== at 0x484DA83: calloc (in /usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so)
==272468== by 0x2ED326: handle_common_semihosting_redirect_command (semihosting_common.c:1931)
==272468== by 0x1FC703: exec_command (command.c:520)
==272468== by 0x1FC703: jim_command_dispatch (command.c:931)
==272468== by 0x36980F: JimInvokeCommand (in /home/marc/openocd/build/src/openocd)
==272468== by 0x1FFFFFFFFF: ???
==272468== by 0x53ED09F: ???
==272468== by 0x400000002: ???
==272468== by 0x1FFEFFF7FF: ???
==272468== by 0x3D3984: ??? (in /home/marc/openocd/build/src/openocd)
==272468== by 0x2: ???
==272468==
Change-Id: I3e5323f145a98d1ff9ea7d03f87ed96140f49a18
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8257
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Currently, errors in pre/post-enable events are ignored and capturing is
always started, even if necessary device configuration fails. This
behaviour is confusing to users. Also, the TPIU must be disabled before
re-configuration is possible.
Start capturing and enable TPIU only if no errors in pre/post-enable
events occurred.
Change-Id: I422033e36ca006e38aa4504d491b7947def1237a
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8254
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
When external capturing is configured (default), the SWO pin frequency
is required. Enforce this to avoid a division by zero error.
While at it, ensure that the 'out_filename' variable always contains a
valid string. This saves a few checks and makes the code more clean and
readable.
Change-Id: If8c1dae9549dd10e2f21d5b896414d47edac9fc2
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8224
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
This would elliminate the need for an extra nop in-between the two reads
in case of a 64-bit register.
Change-Id: I2cddc14f7f78181bbda5f931c4e2289cfb7a6674
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
There was an error introduced by
8319eee9e1.
According to RISC-V Debug Spec 1.0.0-rc1 [3.14.2. Debug Module Contro]:
> 0 (inactive): The module’s state, including authentication mechanism,
takes its reset values (the dmactive bit is the only bit which can be
written to something other than its reset value).
`dmactive` was written together with `hartsel` and `hasel` in
8319eee9e1.
Change-Id: I11fba35cb87f8261c0a4a45e28b2813a5a086078
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
For MPU configs, determine memory access rights
by probing protection TLB. Issuing IHI without execute
permissions can trigger an exception.
No new clang static analyzer warnings.
Change-Id: Iea8eab5c2113df3f954285c3b9a79e96d41aa941
Signed-off-by: Ian Thompson <ianst@cadence.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8080
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Historically, the function cortex_a_dap_write_memap_register_u32()
was used to discriminate the register write in APB-AP CPU debug
against the complex memory access in AHB-AP memory bus.
It has no sense to keep the function and its comment.
Plus, by forcing atomic write it impacts the debug performance.
Drop it!
A further rework to enqueue sequence of atomic writes is needed.
Change-Id: I2f5e9015f0e27fa5a6d8337a1ae25e753e2e1d26
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8231
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Tested-by: jenkins
The command 'cache auto' was introduced with commit cd440bd32a
("add armv7a_cache handlers") in 2015 to allow disabling the cache
handling done automatically by OpenOCD.
This was probably a way to test the cache handling when there were
still the two independent accesses for APB-AP CPU debug and for
AHB-AP memory bus.
The handling of cache for cortex_a is robust and there is no more
reason to disable it.
The command 'cache auto' is not used in any upstream script.
On target aarch64 this command has never been introduced as the
cache is always handled automatically by OpenOCD.
Drop the command 'cache auto' and add it in the deprecated list.
Drop the flag 'auto_cache_enabled' by considering it as true.
Rename the function 'armv7a_cache_auto_flush_all_data()' as
'armv7a_cache_flush_all_data()' and, while there, fix the error
propagation in SMP case.
Change-Id: I0399f1081b08c4929e0795b76f4a686630f41d56
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8230
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
The initial OpenOCD code for Cortex-A (ARMv7a) [1] was merged in
2009 but, due to lack of public documentation for ARMv7a, it was
almost a simple copy/paste from the existing code for Cortex-M
(ARMv7m).
On Cortex-M the same AP provides access to both CPU debug and CPU
memory. This feature is not present on ARMv7a.
To still keep some communality with ARMv7m code, the change [2]
splits the CPU debug access from the CPU memory access by using
two independent AP; this is copied from the system architecture of
TI OMAP3530 which provides to DAP a direct AHB-AP memory bus on
AP#0, separated from AP#1 for the APB-AP CPU debug.
But the direct memory access through the system bus breaks the
coherency between memory and CPU caches, so change [3] added some
cache invalidation to avoid issues.
The code to allow ARMv7a CPU to really read/write in CPU memory
was added by change [4] in 2011. Such still not optimized
implementation was very slow, so it did not replace the access
through the system bus. A selection through DAP's 'apsel" command
was used to select between the two modes.
Only in 2015, with change [5], the speed of CPU read/write was
improved using the DCC_FAST_MODE. But the direct access to the
memory through the system bus remained.
Finally, with change [6] in 2018 the system bus access was dropped
for good, as the new virtual target "mem_ap" could implement such
access in a more clean way.
Only memory access through CPU remained for ARMv7a.
Nevertheless, a useless cache invalidation remained in the code,
decreasing the speed of the write access.
Drop the useless cache invalidate on CPU memory write and the
associated comment, not anymore valid.
Drop the now unused function armv7a_cache_auto_flush_on_write().
This provides a speedup of between 4 and 8, depending on adapter
and JTAG/SWD speed.
Link: [1] 7a93100c2d ("Add minimalist Cortex A8 file")
Link: [2] 1d0b276c9f ("The rest of the Cortex-A8 support from Magnus: ...")
Link: [3] d4e4d65d28 ("Cache invalidation when writing to memory")
Link: [4] 05ab8bdb81 ("cortex_a9: implement read/write memory through APB-AP")
Link: [5] 0228f8e827 ("Cortex A: fix extra memory read and non-word sizes")
Link: [6] fac9be64d9 ("target/cortex_a: remove buggy memory AP accesses")
Change-Id: Ifa3c7ddf2698b2c87037fb48f783844034a7140e
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8229
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Fix possible double free and possible
memory leak while creating an ipdbg hub.
Change-Id: I6254663c27c4f38d46008c4dbff11aa27b84f399
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/8085
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
The gdb subsystem is initialized after the first target examine,
so the field struct target::gdb_service is NULL during examine.
A command "smp off" in the examine event handler causes a SIGSEGV
during OpenOCD startup.
Check for pointer not NULL before dereferencing it.
Change-Id: Id115e28be23a957fef1b97ab66d7273f0ea0dce4
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8216
Tested-by: jenkins
The code for cortex_a allocates the register cache during the very
first examine of the target.
To prevent a segmentation fault in assert_reset(), the call to
register_cache_invalidate() is guarded by target_was_examined().
But for targets with -defer-examine, the target is set as not
examined in handle_target_reset() just before entering in
assert_reset().
This causes registers to not be invalidated while reset a target
examined but with -defer-examine.
Change the condition and invalidate the register cache if it has
been already allocated.
Change-Id: I81ae782ddce07431d5f2c1bea3e2f19dfcd6d1ce
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8215
Tested-by: jenkins
The code for aarch64 allocates the register cache during the very
first examine of the target.
To prevent a segmentation fault in assert_reset(), the call to
register_cache_invalidate() is guarded by target_was_examined().
But for targets with -defer-examine, the target is set as not
examined in handle_target_reset() just before entering in
assert_reset().
This causes registers to not be invalidated while reset a target
examined but with -defer-examine.
Change the condition and invalidate the register cache if it has
been already allocated.
Change-Id: Ie13abb0ae2cc28fc3295d678c4ad1691024eb7b8
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8214
Tested-by: jenkins
Prevent a segmentation fault by preventing to try to halt a target
that has not been examined yet.
Change-Id: I5d344e7fbdb5422f7c5e2c39bdd48cbc6c2a3e58
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8213
Tested-by: jenkins
Commit 16b4b8cf54 ("Cortex-M3: expose most DWT registers") added
the DWT registers to the list of CPU registers.
The commit message from 2009 reports the reason behind this odd
mixing of CPU and DWT registers.
This feature got broken in 2017 with the introduction of the field
struct reg::exist and its further use in the code. As result, the
command 'reg' on a target Cortex-M reports only the core registers
and then the header line
===== Cortex-M DWT registers
not anymore followed by the DWT registers.
Fix it by tagging each DWT registers as existing.
Change-Id: Iab026e7da8d6b8ba052514c3fd3b5cdfe301f330
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: b5964191f0 ("register: support non-existent registers")
Reviewed-on: https://review.openocd.org/c/openocd/+/8198
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
For GDB to fully support hardware watchpoints, OpenOCD needs to tell GDB
which data address has been hit. OpenOCD relies on a target-specific
hit_watchpoint function to do this. If GDB is not given the address, it
will not print the hit variable name or its old and new value.
There does not seem to be a way for the hardware to tell us which
trigger
was hit (0.13 introduced the 'hit bit' but this is optional).
Alternatively,
we can decode the instruction at dpc and find out which memory address
it accesses.
This commit adds support for RVC (compressed) load and store
instructions.
Related to:
https://github.com/riscv-collab/riscv-openocd/issues/688https://github.com/riscv-collab/riscv-openocd/pull/291
This commit is related to testing how OpenOCD responds to `dmi.busy`.
Consider testing on Spike (e.g. `riscv-tests/debug` testsuite). Spike
returns `dmi.busy` if there were less then a given number of RTI cycles
(`required_rti_cycles`) between DR_UPDATE and DR_CAPTURE:
https://github.com/riscv-software-src/riscv-isa-sim/blob/master/riscv/jtag_dtm.cc#L145https://github.com/riscv-software-src/riscv-isa-sim/blob/master/riscv/jtag_dtm.cc#L202
`required_rti_cycles` gets it's value from `--dmi-rti` CLI argument and
is constant throughout the run.
OpenOCD learns this required number of RTI cycles by starting with zero
and increasing it if `dmi.busy` is encountered. So the required number
of RTI cycles is learned during the first DMI access in the `examine()`.
To induce `dmi.busy` on demand `riscv reset_delays <x>` command is
provided. This command initializes `riscv_info::reset_delays_wait`
counter to the provided `<x>` value. The counter is decreased before a
DMI access and when it reaches zero the learned value of RTI cycles
required is reset, so the DMI access results in `dmi.busy`.
Now consider running a batch of accesses. Before the change all the
accesses in the batch had the same number of RIT cycles in between them.
So either:
* Number of accesses in the batch was greater then the value of
`riscv_info::reset_delays_wait` counter and there was no `dmi.busy`
throughout the batch.
* Number of accesses in the batch was less or equal then the value of
`riscv_info::reset_delays_wait` counter and the first access of the
batch resulted in `dmi.busy`.
Therefore it was impossible to encounter `dmi.busy` on any scan of the
batch except the first one.
Change-Id: Ib0714ecaf7d2e11878140d16d9aa6152ff20f1e9
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Recently, (after b503fdef02) OpenOCD started to notify user about hart
state updates. This causes confusion in some cases since some internal
updates to the hart state should not be visible to the user as these are
implementation details. For example situation like this:
```
> reset halt
JTAG tap: riscv.tap tap/device found: 0xdeadbeef ...
> resume
[riscv.cpu0] Found 4 triggers
riscv.cpu0 halted due to single-step.
[riscv.cpu1] Found 4 triggers
riscv.cpu1 halted due to single-step.
[riscv.cpu2] Found 4 triggers
riscv.cpu2 halted due to single-step.
[riscv.cpu3] Found 4 triggers
riscv.cpu3 halted due to single-step.
```
likely confuse people.
There is no issue with the resume functionality. It`s just that
resume internally causes single-step that causes hart state
to change.
This commit disable calling of user-specified (and default)
callbacks during the "hidden" step operation disabling these
confusing messages
Change-Id: I3412a089e2abdcd315d86cec7ee732fdd18c1601
Signed-off-by: Parshintsev Anatoly <anatoly.parshintsev@syntacore.com>
Prior to the commit, pc was cached at `info->dpc`, but dpc at register
cache.
Change-Id: I369788441dbe21bcf8fc360d2e97e98096b25e3a
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
`reg` is a number in register cache, as evident by the following call to
`reg_cache_set()`. `CSR_DCSR` is `GDB_REGNO_DCSR - 65`. This results in
setting cache value for another register, which does not exist, and
causes a segfault if all non-existent registers are not allocated a
value (`reg->value == NULL`).
Change-Id: Iab68a4bb55ce6d4730804e9709e40ab2af8a07c6
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
This allows to merge the implementation in `batch.c` with the one in
`riscv-013.c`.
Change-Id: Ic3821a9ce2d75a7c6e618074679595ddefb14cfc
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
According to the RISC-V Debug Spec (1.0.0-rc1)[3.7 Abstract Commands]:
> While an abstract command is executing (busy in abstractcs is high), a
debugger must not change hartsel, and must not write 1 to haltreq,
resumereq, ackhavereset, setresethaltreq, or clrresethaltreq.
The patch ensures the rule is followed.
Change-Id: Id7d363d9fdeb365181b7058e0ceb0be0df39654f
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
This allows to examine each DM ones (e.g. enumerating harts assigned to
the DM). Additionaly, it is guaranteed that the DM is reset before the
examination.
Change-Id: I2333d06ff1152bf51c647d59baa55cb402054cb9
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
According to the RISC-V Debug Spec (1.0.0-rc1)[3.7 Abstract Commands]:
> While an abstract command is executing (busy in abstractcs is high), a
debugger must not change hartsel, and must not write 1 to haltreq,
resumereq, ackhavereset, setresethaltreq, or clrresethaltreq.
Tracking `abstractcs.busy` allows to enforce this rule.
Change-Id: If5975b48cf9fd379033268145c79103c36fb8134
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Checkpatch complains for extra parenthesis not required.
Drop them.
Change-Id: I311409f5732acf10a4910de5dcf0fb05f43e21b5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8187
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
Initialized `value` variables that could only be set in a branch.
Change-Id: Iec7413ade9d053c93352a58ff954ad49a6545923
Signed-off-by: Walter Ji <walter.ji@oss.cipunited.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8179
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
It was noticed that the remote_bitbang protocol has a design issue:
SWD and sleep commands cannot be implemented at the same time, because
they overlap:
- SWD uses d,e,f,g for setting pin state
- sleep uses d,D for microsecond and millisecond sleep, respectively
This has previously been reported by Marek Vrbka, but it wasn't fixed.
This commit does the following to resolve the issue:
- Change the sleep commands to 'Z' for 1 ms, 'z' for 1 µs
- Document 'D' and 'd' as deprecated aliases
- Switch the remote_bitbang driver in OpenOCD to 'Z' and 'z'
Unfortunately that's a breaking change, because existing adapter-side
implementations of the protocol will have to implement the new commands
to keep working with future versions of OpenOCD. Fortunately, the
remote sleep commands haven't been part of an OpenOCD release yet,
which should limit the breakage somewhat.
Reported-by: Marek Vrbka <marek.vrbka@codasip.com>
Link: https://sourceforge.net/p/openocd/mailman/openocd-devel/thread/670d28d2-75a1-45ec-afe5-541415701d7a%40codasip.com/
Fixes: e8e09b1b5 ("remote_bitbang: add use_remote_sleep option to send delays to remote")
Change-Id: I04d2790a33bff9d47eb7f69b3275fd9a271625ae
Signed-off-by: J. Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-on: https://review.openocd.org/c/openocd/+/8191
Reviewed-by: David Ryskalczyk <david.rysk@gmail.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Reviewed-by: Jeremy Herbert <jeremy.006@gmail.com>
Commit 290eac04b9 ("drivers/linuxgpiod: Migrate to adapter gpio
commands") introduced an incorrect check to determine if the
library libgpiod declares the line request flags:
GPIOD_LINE_REQUEST_FLAG_BIAS_DISABLE
GPIOD_LINE_REQUEST_FLAG_BIAS_PULL_UP
GPIOD_LINE_REQUEST_FLAG_BIAS_PULL_DOWN
The names above are declared by the library inside an enum, thus
cannot be used by the C preprocessor in a #ifdef.
Determine in configure if the version of libgpiod provides the
line request flags for "bias" and define a C macro.
Use the new macro in the driver code.
Change-Id: Iaa452230f4753fce4c6e9daa254299cedb7cab7f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: 290eac04b9 ("drivers/linuxgpiod: Migrate to adapter gpio commands")
Reviewed-on: https://review.openocd.org/c/openocd/+/8186
Tested-by: jenkins
Reviewed-by: Michael Heimpold <michaheimpold@gmail.com>
Move setting of do_reconnect flag from swd_run_inner()
to swd_run(). Reconnect is not used at the inner level
and the flag had to be cleared after swd_run_inner()
to prevent recursion.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: Ib1de80bbdf10d1cbfb1dd351c6a5658e50d12af2
Reviewed-on: https://review.openocd.org/c/openocd/+/8155
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
The file 'list.h', copied from FreeBSD, does not depend from any
OpenOCD specific include file, but only needs 'stddef.h' for the
type 'size_t'.
Let 'list.h' to include the correct header file, then fix the now
broken dependencies in the other files that were incorrectly
relying on 'list.h' to include 'helper/types.h'
Change-Id: Idd31b5bf607e226cac44ef41b2aa335ae4dbf519
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8173
Tested-by: jenkins
The file list.h was originally taken from the Linux kernel code,
thus under license GPL-2.0-only. This locks OpenOCD to follow the
same license, even if the majority of OpenOCD files are licensed
as GPL-2.0-or-later.
A similar file is also present in FreeBSD code base under the more
permissive license BSD-2-Clause.
Drop the code from Linux kernel and replace it with the code from
FreeBSD 13.3.0.
Adapt the code to OpenOCD coding style by fixing the majority of
issues identified by checkpatch.
Add the OpenOCD specific macros and comments.
Unfortunately this causes the lost of all the doxygen comments.
Checkpatch-ignore: MACRO_ARG_REUSE, MACRO_ARG_PRECEDENCE
Change-Id: I6d86752c50158f3174c4e8c4add81e9998d01e0e
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8172
Tested-by: jenkins
OpenOCD can send it's log to gdb, and gdb replies with 'OK'.
Calls to LOG_XXX() are also present in the code that communicates
with gdb. This can cause infinite nested calls.
OpenOCD uses the flag 'gdb_con->busy' to protect the communication
with gdb and prevent nested calls.
There is no reason to check for 'gdb_con->busy' in the code for
keep-alive, as keep_alive() is never called in this gdb server;
the flag would eventually be set if the current keep_alive() will
send something to gdb.
Drop the flag 'gdb_con->busy' in gdb_keep_client_alive().
While there, document the use of 'gdb_con->busy'.
Change-Id: I1ea20bf96abb5d2f1fcdba1e3861df257c396bb6
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8166
Tested-by: jenkins
To avoid gdb to timeout, OpenOCD implements a keep-alive mechanism
that consists in sending periodically to gdb empty strings embedded
in the "O" remote reply packet.
The main purpose of "O" packets is to forward in the gdb console
the output of the remote execution; the gdb-remote puts in the "O"
packet the string that gdb will print. It's use is restricted to
few "running/execution" contexts listed in
http://sourceware.org/gdb/onlinedocs/gdb/Stop-Reply-Packets.html
and this currently limits the keep-alive capabilities of OpenOCD.
Long data transfer (memory R/W) can also cause gdb to timeout if
the interface is too slow. In this case the usual keep-alive based
on "O" packet cannot be used and, if used, would trigger a protocol
error that causes the transfer to be dropped.
The slow transfer rate can be simulated by adding some delay in the
main loop of mem_ap_write() and mem_ap_read(), then using the gdb
commands "dump" and "restore".
In the wait loop during a memory R/W, gdb drops any extra character
received from the gdb-remote that is not recognized as a valid
reply to the memory command. Every dropped character re-initializes
the timeout counter and could be used as keep-alive.
From gdb 7.0 (released 2009-10-06), an asynchronous notification
can also be received from gdb-remote during a memory R/W and has
the effect to reset the timeout counter, thus can be used as
keep-alive.
The notification would be treated as "junk" extra characters by any
gdb older than 7.0, being still valid as keep-alive.
Check putpkt_binary() and getpkt_sane() in gdb commit
74531fed1f2d662debc2c209b8b3faddceb55960
Currently, only one notification packet ("Stop") is recognized by
gdb, and gdb documentation reports that notification packets that
are not recognized should be silently dropped.
Use 'set debug remote 1' in gdb to dump the received notifications
and the junk extra characters.
Add a new level in enum gdb_output_flag for using the asynchronous
notifications.
Activate this new level during memory transfers.
Send a custom "oocd_keepalive" notification packet as keep_alive.
While there, drop a useless return in the switch/case, already
managed in case of break.
After this commit, the proper calls to keep_alive() have to be
added in the loops that code the memory transfers. Of course, the
keep_alive() should be placed during the wait for JTAG flush, not
while locally queuing the JTAG elementary transfers.
Change-Id: I9ca8e78630611597d15984bd0e8634c8fc3c32b9
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8165
Tested-by: jenkins
The logic in `init_registers()` was quite convoluted.
Initialization of each `struct reg` field is separated into function
`gdb_regno_<field_name>()`.
IMHO, this makes it much easier to reason about the code.
Change-Id: Id7faa1464ce026cc5025585d0a6a95a01fb39cee
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Add access to fpr and cp1 registers.
GDB can now check the FPRs with `info reg f` and change them.
Checkpatch-ignore: MACRO_ARG_REUSE
Change-Id: I63896ab6f6737054d8108db105a13a58e1446fbc
Signed-off-by: Walter Ji <walter.ji@oss.cipunited.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7866
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
To simplify the ipdbg start/stop command and be able to
add additional commands in the future, we introduce the
concept of a hub which has to be created before a
ipdbg server can be started.
The hub was created on the fly in previous versions.
Change-Id: I55f317542d01a7324990b2cacd496a41fa5ff875
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/7979
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
When receiving an OpenOCD debug log to investigate about errors or
issues, the first question is often about providing the complete
command line to better understand the use context.
Plus, when OpenOCD is lunched by an IDE, its command line is kept
hidden inside the IDE, adding troubles to the user to recover it.
Add the full command line directly inside the debug log.
It could have been useful to also search and add in the log the
full path of the OpenOCD executable, but this is not an immediate
task due to portability among OS's. See, for example:
https://stackoverflow.com/questions/933850
This part could be handled in a future change, if really needed.
Change-Id: Ia6c5b838b9b7208bf1ecac7f95b5efc319aeabf5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8170
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Extend the existing code to support Monitor mode in AArch32.
Change-Id: Ia43df98d1497baac48aea67b92d81344c24f0635
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8169
Tested-by: jenkins
ADIv6 brought more complicated rules for DP reg 0 banking.
Neither the original implementation [1] nor the later
modification [2] respected that the DP reg 0 is banked
for read only, not for write. Enforcing of an useless
SELECT write before a write to ABORT register may trigger
FAULT (CTRL/STAT bits ORUNDETECT and STICKYORUN are set)
or WAIT (DP is stalled by an outstanding previous operation)
and therefore make ABORT register virtually unusable
on some adapters (bitbang, CMSIS-DAP).
There are DP ABORT specific functions swd_queue_ap_abort()
and swd_clear_sticky_errors() which worked around the problem
using the lowest level swd->write_reg(). Using a specific
write procedure for a single DP register was error prone
(there are other DP_ABORT writes using swd_queue_dp_write_inner())
and also the Tcl command 'xx.dap dpreg 0 value' suffered
from unwanted SELECT write.
Other smaller discords in DP banking probably do not
influence normal DP operation however they may complicate
debugging in corner cases.
Adhere strictly to the DP banking rules for both ADI versions.
Fixes: [1] commit 72fb88613f ("adiv6: add low level swd transport")
Fixes: [2] commit ee3fb5a0ea ("target/arm_adi_v5: fix DP SELECT logic")
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I3328748c1c3e0661c5ecd6eb070ac519b190ace2
Reviewed-on: https://review.openocd.org/c/openocd/+/8154
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
The bitbang driver kept retrying a SWD command as long as
the debugged device had been responding by SWD WAIT.
If the DP stalled in WAIT permanently, OpenOCD hanged.
Check 0.5 sec timeout in WAIT retry loop.
While on it insert a short alive_sleep() if the command
is retried 20 or more times.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I744e56e21a5a2dc2c4494cc0d7bbcb4be14ddb23
Reviewed-on: https://review.openocd.org/c/openocd/+/8153
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Log SWD commands with not OK response but WAIT retries at debug level.
For commands responded OK and WAIT retries use debug io level
not to flood the log.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: Idf658e82ed970061c155945df55d06908ed25e09
Reviewed-on: https://review.openocd.org/c/openocd/+/8152
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Allow logging at a changeable level.
Add an example of usage in ftdi driver.
Log SWD commands with not OK response at debug level (3).
For commands which responded OK use debug io level (4)
not to flood the log.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I67a472b293f7ed9ee84cadb7c081803e9eeb1ad0
Reviewed-on: https://review.openocd.org/c/openocd/+/8151
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Allow direct pointer to struct adiv5_private_config
for targets with adiv5_private_config inside of a bigger
private config container. Use it instead of the private_config
pointer toggling hack in aarch64.c
Allow optional use of -dap parameter and use it instead
of the static variable hack in xtensa_chip.c
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I7260c79332940adfa49d57b45cae39325cdaf432
Reviewed-on: https://review.openocd.org/c/openocd/+/8138
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Unify the error codes returned by adapter drivers in the case
of the received SWD ACK field differs from OK.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I29e478390b4b30408054a090ac6a7fac3415ae71
Reviewed-on: https://review.openocd.org/c/openocd/+/8137
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Commit [1] added a break on error to the nrf5_erase() sector loop
and the checking of the res value became useless in the for loop condition.
Removing nrf5_get_probed_chip_if_halted() later in [2]
dropped res initialization and clang static analyser complains
"The left operand of '==' is a garbage value"
Drop the useless test!
Fixes: [1] commit 491636c8b8 ("flash/nor/nrf5: check protection before flash erase/write on nRF51")
Fixes: [2] commit 2db325f539 ("flash/nor/nrf5: drop nrf5_get_probed_chip_if_halted()")
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: Ife6071c509719f8d7dc312fe9a780bdcf2575f69
Reviewed-on: https://review.openocd.org/c/openocd/+/8174
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Add access to dsp registers and a command for dsp related operations.
Checkpatch-ignore: MACRO_ARG_REUSE
Change-Id: I30aec0b9e4984896965edb1663f74216ad41101e
Signed-off-by: Walter Ji <walter.ji@oss.cipunited.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7867
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Running the GDB command 'flash-erase' triggers sending the remote
GDB commands 'vFlashErase' (one per flash bank) followed by one
single 'vFlashDone', with no 'vFlashWrite' commands in between.
This causes the field 'gdb_connection->vflash_image' to be NULL
during the execution of 'vFlashDone', triggering a segmentation
fault in OpenOCD.
While parsing 'vFlashDone', check if any image to flash has been
received.
Change-Id: I443021c7a531255b60f2c44c2685e52e3c34b5c8
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8164
Tested-by: jenkins
After commit d9b2607ca0 ("gdb_server: support sparse register
maps"), GDB crashes while requesting the value of 'cpsr' because
the fake register is tagged as not existing.
Change the logic and set all register as existing, while still
limiting the list for the initial GDB request at connect.
Change-Id: I1c4e274c06147683db2a59a8920ae5ccd863e15c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8163
Tested-by: jenkins
for features flags.
Change-Id: I8bff0f5fac41c50180c847f36c6d2a075eca32ca
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8109
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
The command substantially complicates support of nRF53/91
series. It was not even properly ported to nRF52.
The informative value is disputable. Who wants to see
e.g. override trim values for radio or unique device ID?
Drop it and simplify the driver.
Change-Id: Ia7fb20ce2ebf16065705c5d18deaf934e58db426
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8108
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
nrf5_get_probed_chip_if_halted() was somewhat bizarre
combination of functions:
- test if the target is halted is appropriate for flash
erase/write only, certainly not for getting chip info
- getting chip pointer takes place more frequently
and using one temporary variable for dereference
makes no harm
- probing chip is useless at all as the flash
infrastructure always calls auto_probe() before
entering a flash operation
Replace the function by ordinary and readable code.
Change-Id: Ic31f4e33d8b7b36687be3f40bfd0fe913d17b75f
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8107
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
from Product Specification of nRF52805, 810, 811
820, 833 and 840.
While on it, rename the table to make sure the codes
are valid for nRF52 series only.
Change-Id: Id8f78fd214c5d345d1769378ae546a6be5a183ba
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8105
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Use the newer driver name 'nrf5' instead.
While on it set the unused parameters of flash bank
creation to zero.
While on it remove 2 empty comments.
Change-Id: I9cf0eadc5b696e6c8b7e6aec0ea3345967523e87
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8103
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
While on it update table comment and drop
not working URLs.
Change-Id: I9e21c72aa75a908c644460e43c148d3240c49b2d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8102
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
On 2008-03-05, before git's age, commit 6d95014674 adds sending
an ACK ('+' char) at GDB connection, before receiving any GDB
remote command that requires to be ACK'ed.
Neither the text added in the commit message ("added ACK upon
connection (send +)") nor in the associated comment ("send ACK to
GDB for debug request") provide an exhaustive explanation for
sending this unsolicited ACK.
This code has never been touched since its introduction.
Analysis of GDB code doesn't show it's required, including old GDB
code.
Running gdbserver (from GDB package) and attaching it with "nc"
shows that gdbserver does not send any ACK to a new connection.
Same for lldb-server.
Drop it!
Change-Id: Id68c352ce44dd85a1ea3d67446e17e2a241ef058
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/6768
Tested-by: jenkins
Reviewed-by: Jan Matyas <jan.matyas@codasip.com>
Reviewed-by: Anatoly P <anatoly.parshintsev@syntacore.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
For the sake of https://review.openocd.org/c/openocd/+/7957
Instead of "coreid", 'target smp' command call order used as
an index
Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: Iab86b81868d37c0bf8663707ee11367c41f6b96d
Reviewed-on: https://review.openocd.org/c/openocd/+/8162
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
When performing a command queue allocation larger than the default page
size of 1MiB any subsequent allocations will run into an integer under-
flow when checking for the remaining memory left in the current page.
Causing the function returning a pointer past the end of the buffer and
thus creating a buffer overflow.
This has been observed to cause some transfers to Efinix FPGAs to fail,
because another buffer can get corrupted in the process, causing its
respective free() to fail.
Change-Id: Ic5a0e1774e2dbd58f1a05127f14816c8251a7d9c
Signed-off-by: SydMontague <sydmontague@phoenix-staffel.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/8126
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
find members of a union, nested in struct.
Allows file to be compiled with GCC 4.0
Signed-off-by: Sevan Janiyan <venture37@geeklan.co.uk>
Change-Id: Ied68668d3b5f811573a20e11e83aceff268963eb
Reviewed-on: https://review.openocd.org/c/openocd/+/8120
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
At least one TAP shouldn't be in BYPASS mode
Change-Id: Ic882acbfc9b6a9f4b0c3bb4741a49f3981503c8c
Signed-off-by: Kirill Radkin <kirill.radkin@syntacore.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7741
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Fix flash operation error when addr-width > 32bit
on any 32-bit OS and some 64-bit OS (windows).
Change-Id: I199f1cc5128c45bd0bb155e37acb2fb6325dff88
Signed-off-by: wangyanwen <wangyanwen@nucleisys.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8095
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
Allows build with legacy toolchains which do not support
C23 nor GCC extension for binary literals.
Change-Id: I742d3a8a86bf16f81421d11c59d3cb155ee17aed
Signed-off-by: Sevan Janiyan <venture37@geeklan.co.uk>
Reviewed-on: https://review.openocd.org/c/openocd/+/8123
Tested-by: jenkins
Reviewed-by: Jörg Wunsch <openocd@uriah.heep.sax.de>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Add more debug messages connected with triggers.
Update names for internal flags to make them more clarified.
Change-Id: I5642346ce4a1e9bf79b22cdbf36bd757a7beffa8
Signed-off-by: Kirill Radkin <kirill.radkin@syntacore.com>
Evgeniy Naydanov
Evgeniy Naydanov
Tomas Vanek
Antonio Borneo
Marc Schink
Timur Golubovich
Daniel Anselmi
Paul Fertser
Parshintsev Anatoly
Mark Featherston
Noah Moroze
Anatoly Parshintsev
Jan Matyas
Bernhard Rosenkränzer
Ian Thompson
Aleksey Lotosh
Walter Ji
J. Neuschäfer
Emmanuel Blot
Steven Chang
Erhan Kurubas
Luca Rufer
SydMontague
Sevan Janiyan
Kirill Radkin
wangyanwen